May, Friday 4th
11h (room 2014, 'Digiteo Shannon' 660 building) (
see location)
Dominique Fourer
(IRCAM)
Abstract:
Real-world signals are often mixtures of elementary components (sinusoids, transients, noise) that require the best trade-off between accuracy and robustness of the analysis to be efficiently exploited. In this presentation, we present recent advances in the field of signal processing theory for disentangling meaningful components from an observed mixture. We illustrate the proposed methods with practical applications (not limited) to music information retrieval.
Contact: guillaume.charpiat at inria.fr
All TAU seminars:
here